AI’s biggest safety debate is widening, and the sharpest disagreement is no longer hidden behind technical language. Nick Clegg has dismissed fears that artificial intelligence holds “godlike power to exterminate humanity,” while other technology leaders warn that poor decisions could still place people at risk.
The clash reaches from Silicon Valley to the United Nations, raising a central question: should policymakers focus on extreme extinction scenarios, or build firm rules around the risks already taking shape?
Clegg Challenges the Extinction Narrative
Nick Clegg, the former UK deputy prime minister and former president of global affairs at Meta, said technology leaders should focus on cybersecurity and bioweapons threats instead of “slightly hand-wavy” views of AI extermination. He described the current discourse on AI risks as “totally silly,” arguing that engineers can become trapped inside a culture that magnifies the most dramatic predictions.
“There is a culture in Silicon Valley – it’s an odd place for those who haven’t been there,” Clegg said. “It is in many ways the centre of the universe, or at least claiming to shape it. It’s also an oddly parochial place.”
He said engineers can get close to the technology, “start slightly breathing their own fumes” and “wind themselves up into a lather” about AI risks. Clegg also mentioned that one person who runs an AI lab said the current discourse was “totally silly,” adding another voice to the argument that extinction fears can distort public debate.
Andrew Ng, an AI pioneer, has taken an even harder line. Ng called fears of AI causing human extinction “science fiction” and said warnings about existential AI risks from researchers at top model makers may hurt efforts to deliver the technology’s maximum benefits to society. He also said fears about catastrophic AI harms had previously been used to shape regulation and publicity.
Other Leaders Urge Caution and Regulation
Dario Amodei, CEO of Anthropic, offered a different warning. “If managed poorly, I even believe AI could be a risk to humanity as a whole,” he said. That position does not claim that extinction is inevitable, but it places responsibility on the companies and institutions building and deploying advanced models.
The United Nations has become a stage for that concern. At the UN, heads of major artificial intelligence firms pleaded with the organization to regulate the fast-expanding technology they have been designing, seeking action to prevent risks to humanity.
Clegg does not reject every form of AI oversight. He said there is a case for “tramlines and greater transparency” around autonomous AI hacking incidents, pointing toward clear limits for systems that can act without direct human control.
He also argued that transparency and regulation could improve safety across the AI industry. “There is a lot that can be done both voluntarily and through regulation, to enforce far greater transparency on the way in which these models are concocted and how they operate,” Clegg said.
That approach would shift attention from predictions about machine extinction toward practical questions about how models are built, how they operate and what happens when autonomous systems take action.
Why Clegg Rejects a Single Control Point
Clegg opposed the idea of a “kill switch” for AI and rejected producing only closed models, citing safety concerns. He argued that concentrating control among a small group of companies could create its own danger by limiting scrutiny and narrowing the political debate.
“I think it is perhaps accidentally self-serving, because it sort of implies that we can only be made safe by having everything controlled, in effect, by an oligopoly of a small number of companies, and I think it slightly paralyses the political debate,” he said.
His position supports a model where regulation and openness work together, rather than placing every safety decision inside a small number of private organizations. For Clegg, transparency is not a replacement for responsibility; it is one of the tools that can make responsibility possible.
The debate also reaches beyond the United States. Clegg suggested collaboration among India, Europe and the US on AI development, calling them “the big three techno democracies in the world.” He said, “I think there’s a great potential for the big three techno democracies in the world – India, Europe and the US – to collaborate more.”
He also warned that the UK is not a world leader on AI and stressed the need for realistic influence. “We need to be realistic about what influence we can and cannot exert,” Clegg said.
A Debate Shaped by Interests and Consequences
Clegg’s view also sits within a personal connection to the AI industry. He is a cofounder of AI startup AMI Labs and owns more than 917,000 shares in Nscale, details that form part of the wider context around his comments.
That context does not settle the argument, but it shows why AI safety debates attract close attention. Industry leaders are discussing rules for systems connected to their companies, their research and their future influence, while governments are deciding how much authority those systems should receive.
One side sees extinction warnings as science fiction that can distract from real threats such as cybersecurity and bioweapons. Another warns that AI could become a risk to humanity if people manage it poorly. Between those positions sits a growing push for transparency, regulation, cooperation and clear limits on autonomous action.
The next stage of AI policy will test whether those ideas can move beyond speeches. India, Europe, the US, the UK and the United Nations now face the same challenge: support AI development while creating enough visibility and control to prevent the technology from outrunning public safety.
Based on
